FR2764708A1 - IMPROVEMENTS TO REAL-TIME RADIO-SATELLITE LOCATION METHODS AND SYSTEMS, IN PARTICULAR OF THE GPS TYPE - Google Patents

Abstract

Method for locating points on the surface of the earth in which one simultaneously picks up at a reference station of known coordinates and on a mobile which is moved to the points to be located, radiofrequency signals emitted on two frequencies (L1 and L2) in the direction of the earth via a constellation of satellites, and an initialization processing is implemented by which phase ambiguity is removed from the signals coming from the various satellites observed, as well as a kinematic processing allowing, at the end of the initialization processing, of following the phase turns of the signals while the mobile is moving, characterized in that the initialization processing implements a determination of coordinates of points from combinations of measurements of phases corresponding to two apparent frequencies (L4 and L5) of wavelengths greater than those of the emission frequencies (L1 and L2) of the constellation of satellites and in that the measurements resulting from the removal of ambiguities carried out at each of these two frequencies are linearly combined so that their noises are canceled out.

Description

The present invention relates to radio localization in

real time from point to point

the surface of the globe.

  The method and the system proposed by the invention advantageously find application in all the techniques requiring knowledge of the position of a mobile on the surface of the globe, in three dimensions, with centimetric precision. Let us quote for example: geodesy, topography, seismic, hydrography,

robotics ...

  For a general presentation of radio localization techniques and in particular GPS techniques (Global Positioning System), one can advantageously refer to the introduction to the patent.

  Applicant's French published under No. 2,715,230.

  Remember that the GPS system transmits on two

  frequencies generally called L1 and L2.

  The applicant had proposed in the aforementioned patent a radio localization technique allowing an operator to carry out acquisitions with centimetric precision, without it being necessary for him to carry out initialization by static determination before the progress of his kinematic reading. , nor to return to calibration points, in the event of loss of satellite signals.

  The algorithms mainly described correspond to

  lay at the operation by the system of the only frequency L1. The initialization phase incorporated a rapid convergence process, possibly implemented while the operator and his receiving antenna were moving - which made it possible to start kinematic monitoring after a period of the order of 3 to 10 minutes. . The durations of 3 to 10 minutes indicated in the aforementioned patent therefore correspond to a monofrequency operation. The invention described here proposes an extension of this principle which, by using in a completely original way the two frequencies of the system, makes it possible to reduce to a few seconds the initialization times, in particular, but not limitatively, when one implements a convergence algorithm of the

  same type as that described in the aforementioned patent.

  It is recalled that the emission frequencies L1, L2 of the satellites of the GPS constellation (NAVSTAR) are respectively equal to 1.575.42 GHz and 1.227.60 GHz and correspond to wavelengths of 19 cm and

24.4 cm.

  To facilitate the search for ambiguities, it is conventionally proposed to use, not directly the phase measurements made at frequencies L1 and L2, but a linear combination of these measurements corresponding to the frequency L3 = L1 - L2 (technique called "WideLaning "according to Anglo-Saxon terminology and well known to those skilled in the art). This technique offers the advantage of a much wider ambiguity (86 cm against 19 cm for L1), but at the expense of an increase in noise in an even greater ratio (1.9 cm assuming a

  0.3 cm noise on L1 and L2 measurements).

  The invention proposes to use linear combinations of the basic frequencies L1 and L2 giving even much wider ambiguities, with a treatment making it possible to compensate for the very high noises generated.

by these combinations.

  The invention therefore proposes a method for locating points on the surface of the earth in which radio frequency signals transmitted are simultaneously picked up at a reference station of known coordinates and on a mobile which is moved to the points to be located. on two frequencies (Li and L2) towards the earth by a constellation of satellites, and an initialization processing is implemented by which the ambiguity of the phase is removed on the signals coming from the various satellites observed, thus a kinematic processing allowing, at the end of the initialization processing, to follow the phase turns of the signals while the mobile is moving, characterized in that the initialization processing implements a determination of point coordinates from linear combinations of phase measurements corresponding to two apparent frequencies of wavelength greater than those of the emitting frequencies on (L1 and L2) of the constellation of satellites and in that one linearly combines the positions resulting from the ambiguity surveys carried out at each of these two frequencies of

so that their noises cancel each other out.

  As will be understood, ambiguity surveys are thus carried out on signals whose frequencies correspond to long wavelengths, but the effects of noise at these frequencies are attenuated, so

  that we have a reliable determination.

  In the application of the general principle to the GPS system, the phase combinations preferentially correspond to the frequencies L4 = 4L2-3L1 and L5 =

4L1 - 5L2.

  The noises generated by these combinations are of opposite sign and in the ratio 1.43, so that they are compensated for by the following linear combinations:

X = (1.43 XL4 + XL5) / 2.43

Y = (1.43 YL4 + YL5) / 2.43

Z = (1.43 ZL4 + ZL5) / 2.43

  o XL4, YL4, ZL4 and XL5, YL5, ZL5 are the coordinates determined for this point from phase measurements

at frequencies L4 and L5.

  Furthermore, the validation of the final result is considerably improved by making a comparison between, on the one hand the evolution of the phases measured between an instant to corresponding to the start of the treatment and the current instant tl, and on the other hand the evolution of the Mobile-Satellite distances calculated between these same instants from the point to be validated. Compared to the residue evolution test described in FR 2 715 230, this test has the advantage of being performed on the entire phase measurement and not only on its fractional part. This results in a very

  important sensitivity and therefore reliability.

  Furthermore, in order to fully exploit the possibilities offered by the use of the two frequencies L1 and L2, a correction process for the ionospheric delay is implemented. This processing applies in all the phases of computation: initialization and kinematics. In general, we exploit the fact that the ionospheric delays observed on the two frequencies L1 and L2 are inversely proportional to the square of the frequencies. In this way, the effect of this delay is compensated for by applying to the calculated positions the following relationships:

X = (1.65XL1-XL2) / 0.65

Y = (1.65YL1-YL2) / 0.65

Z = (1.65ZL1-ZL2) / 0.65.

  o XL1, YL1, ZL1 and XL2, YL2, ZL2 are the coordinates calculated from the phase measurements made at

frequencies Li and L2.

  The same principle is applicable to the calculation of residues. The originality of this calculation is that it applies

  on results, not on measures.

  Other features and benefits of

  the invention will become more apparent from the description which

  follows. This description is purely illustrative and not

  limiting. It should be read with reference to the appended drawings in which: - Figure 1 is a schematic representation of a system according to an embodiment of the invention; - Figure 2 is a block diagram representation of the processing performed by the method according to a

  possible mode of implementation of the invention.

  The location system which has been shown in FIG. 1 mainly comprises a fixed station 1 placed at a point of known coordinates and a mobile 2 which an operator moves to the points which it is desired to locate. The station 1 and the mobile 2 are each provided with means 3 for the reception of the signals transmitted by the GPS satellites of the constellation NAVSTAR. These means 3 are conventionally known to those skilled in the art and in particular comprise a GPS reception antenna, as well

  only means for signal demodulation.

  The base station 1 also includes means 4 for radiofrequency transmitting information

for mobile 2.

  The mobile 2 is provided with receiving means 5 for receiving the information sent by the means

4 from station 1.

  The reception means 3 of the station 1 and of the mobile 2 record the carrier phase of the signals received from each satellite observed. The data thus received by station 1 are transmitted to the

  mobile 2 by the transmission means 4.

  The base station 1 does not carry out any calculation and is content to retransmit the signals which it receives at the

mobile 2.

  The phase differences between, on the one hand, the signals received at the base station 1 and, on the other hand, the signals received at the mobile 2, are treated, as will now be described by a unit U which presents the mobile 2, so as to remove the phase ambiguities and thus determine the position of the mobile 2, knowing the position of the base station 1 as well as the

  ephemeris of the satellites observed.

  The mobile 2 has display means 2a to which the calculated coordinates are transmitted. It also includes means for memorizing these coordinates as well as a keyboard (not shown) allowing the operator, when the antenna has been placed on a point to be located, to control the storage in

  memory of the calculated coordinates displayed.

  The processing carried out by unit U takes place

  according to the flowchart shown in Figure 2.

  In a first step (step 6), the unit U

  checks the quality of the radio signals received.

  When there is not enough information,

  a message 7 is displayed on the means 2a of the mobile 2.

  If on the other hand the information is sufficient, the unit U implements an initialization processing

  fast 8, then kinematic follow-up processing 9.

  The processing 8 comprises three main steps consisting: - in determining an approximate unambiguous position (step 10); - in a registration process on a precise position (step 11);

  - in validity tests (step 12).

  In step 10, a process for determining an unambiguous approximate position is implemented, by combining the use of triple phase differences at frequencies Li and L2 and that of the pseudodistances extracted from the measurements made on the "P" codes. , well known to those skilled in the art. This step corrects the results of the ionospheric correction effect according to

the process described above.

  In step 11, the rapid convergence registration process is implemented corresponding to the treatments described in the patent of the Applicant FR-

  2-715-230, which can advantageously be referred to.

  However, the process is first implemented, no longer on the frequency L1 as described in the aforementioned patent, but by using linear combinations of the frequencies L1 and L2 giving two apparent frequencies L4 and L5 of much longer wavelength . The linear combinations chosen, 4L2 - 3L1 and 4L1 - 5L2, correspond to wavelengths of 1.64 m for L4 and 1.83 m for L5, which generate an amplification of noise of 12.9 cm on L4, 18 , 5 cm on L5, with an assumption of 3 mm on L1 and 4 mm on L2. Instead of directly using the results of positions L4 and L5 marred by these very high noises, a specific linear combination is used which makes it possible to reduce to

almost completely noise.

  With the aforementioned values for L4 and L5, the noises generated on the phase measurements on L4 and on

  L5 have opposite signs, and in the ratio 1.43.

  If one calls XL4, YL4, ZL4 and XL5, YL5, ZL5 the coordinates of the points determined on the phase measurements on L4 and L5, a resulting point without bias is given by:

X = (1.43 XL4 + XL5) / 2.43

Y = (1.43 YL4 + YL5) / 2.43

Z = (1.43 ZL4 + ZL5) / 2.43

  The processing which has just been described makes it possible to initialize the position calculated for the mobile in a very rapid time (from a few seconds to tens of

  seconds for 99.9% reliability).

  The values given previously for the frequencies L4 and L5 are values suitable for the GPS system, however other values may well

heard be considered.

  At the end of this initialization phase, a recalibrated point is determined by implementing the technique of correcting the effect of ionospheric errors. We know that the ionospheric delay is inversely proportional to the square of the working frequency. As a result, the ionospheric delays between two frequencies are linked in a given ratio, which is for example 1.65 for the delays between L2 and L1. At great distances (i.e. greater than km), it is generally the ionospheric delay which is

the main cause of the error.

  When this is the case, the observation of the residual errors of the calculations made at two frequencies, once the ambiguities on these two frequencies have been removed, shows a proportionality of the deviations. For example:

6XL2 = 1.65.6XL1

ÈYL2 = 1,65.5YL1

ÈZL2 = 1.65.8ZL1

  In addition, the residues observed on each

  satellite are in the same report. Due to the non

  correction of ionospheric errors upstream, these residues can be high, to the point of disturbing the consistency tests. But when the residues are high due to the ionospheric delay, the composite residue Res = (, 65 * ResLl-ResL2) / 0.65 o ResL1 and ResL2 are respectively the residues on Ll

and L2, is effectively zero.

  The residue tests of step 15 or implemented during kinematic determinations are therefore preferably implemented with such composite residues, which can be used for consistency tests of position lines, even during strong ionospheric disturbances. The implementation of this technique allows the operation of the system at distances far greater than 10 km usually

  considered a maximum limit.

  At the end of the determination of the approached points, then readjusted by the treatments of stages 10 and 11, one implements in stage 12 tests of validity, and in particular a test 13 of control of the residues (also called test of consistency) similar to that described in FR 2 715 230, as well as a test 14 on the comparative evolution of the measurements and the distances

  calculated from the position to be validated.

  In this test 14, a comparison is implemented between on the one hand the evolution of the phases measured at an instant t0 corresponding to the start of processing and a current instant tl and on the other hand the evolution of the theoretical mobile-satellite distances between these same moments. The identity between the two developments is not

effect verified only to the point.

  Thus, the validity test takes into account not only the fractional parts of the transmission phases but also their number of turns, which makes it much more sensitive and also makes it possible to detect the jumps of turns which could have intervened on the either of the distances measured between

  the initial instant and the current instant.

  During the kinematic phase (step 9), we

  calculates a point using only the frequencies L1 and L2.

  The validation tests of the position of the kinematic follow-up repeating, at each calculation step, those of step 12 as regards the tests on the residue controls and the control tests

phase evolution.

  Also, in this kinematic phase, the determination of the position coordinates made on L1 and L2 is obtained by calculating the coordinates X, Y, Z from the relationships:

X = (1.65 XL1-XL2) / 0.65

Y = (1.65YL1-YL2) / 0.65

  z = (1.65ZL1-ZL2) / 0.65, thus taking into account the effect of ionospheric errors. In these expressions XL1, YL1, ZL1 and XL2, YL2, ZL2 are the coordinates determined on the

  phase measurements observed on L1 and L2.

Claims (6)

  1. A method for locating points on the surface of the earth in which radio frequency signals emitted on two frequencies are simultaneously picked up at a reference station of known coordinates and on a mobile which is moved to the points to be located. L1 and L2) towards the earth by a constellation of satellites, and an initialization processing is carried out by which the phase ambiguity is removed from the signals from the different satellites observed, as well as a processing kinematics allowing, at the end of the initialization processing, to follow the phase turns of the signals while the mobile is moving, characterized in that the initialization processing implements a determination of point coordinates from combinations linear phase measurements corresponding to two apparent frequencies (L4 and L5) of longer wavelengths than those of the emission frequencies (L1 and L2) of the constellation of satellites and in that one linearly combines the positions resulting from the ambiguity surveys carried out at each of these two frequencies so that their noises vanish.   2. GPS radiolocation method according to claim 1, characterized in that the phase measurements are performed at frequencies L4 = 4L2-3L1 and L5 = 4L1-5L2, where L1 and L2 are the two GPS transmission frequencies, and in what we calculate the noiseless coordinates X, Y, Z of a point by: X = (1.43 XL4 + XL5) / 2.43 Y = (1.43 YL4 + YL5) / 2.43 Z = (1.43 ZL4 + ZL5) / 2.43   o XL4, YL4, ZL4 and XL5, YL5, ZL5 are the coordinates determined for this point from phase measurements at frequencies L4 and L5.   3. Method according to one of claims   above, characterized in that during the initialization processing or during the kinematic processing, a position validity test is carried out which performs a comparison between on the one hand the evolution of the phases measured between an instant t0 corresponding to the start of treatment and instant tl current and other   share the evolution of the theoretical mobile distances-   satellites calculated between these same moments since the point to validate.   4. Method according to one of claims   above, characterized in that during the initialization processing and during the kinematic processing, a delay correction processing is implemented   ionospheric applied to the calculated positions.   5. Method according to claim 4, characterized in that one calculates residues ResLi, ResLj on phase measurements at two frequencies Li and Lj and in that one also calculates a composite residue: Res = (X * ResLi- ResLj) / (X-1) o X is the ratio of ionospheric delays between the frequencies Lj and Li.   6. Method according to claim 4, characterized in that during kinematic monitoring, the coordinates X, Y, Z of a point are determined by calculating: X = (1.65XL1-XL2) / 0.65 Y = (1.65YL1-YL2) / 0.65 Z = (1.65ZL1-ZL2) / 0.65.   o XL1, YL1, ZL1 and XL2, YL2, ZL2 are the coordinates corresponding to the phase measurements carried out at frequencies L1 and L2.